84 FLORA INDICA. [Schizandracea.
antheras coalita. Anthem lincari-oblongee, connective lateraliter adnatee, longitudi-
naliter dchiscentes. Bacca in capitulurn globosum diam. 1-2-poU, congests, cu-
neato-suhglobosrc, coccmesc, pisi vel i'abte minoris magn. Semina 1-2.
The sweet but flavourless fruit,of this species is eaten by the inhabitants of Sik-
kini and Eastern Bengal. Though confounded by Wallich with K. Japonica, it ap-
peal's to be quite distinct from the plant figured and described by Siebold and Zuc-
cariiii. We have not seen Japanese specimens, but a specimen in the Hookerian
Herbarium, brought from Hongkong by Major Champion, and referred by Beutham
without hesitation to JT. Japonica, has larger flowers on very short pedicels, which are
more covered with bracts than those of the Indian plant. ' The leaves are also
thicker and firmer, scarcely toothed, and Jonger-petioled. The leaves vary much in
shape in all the species, and, as is often the case among scandent plants, the foliage of
the long suckers is very different from that of the lateral shoots of the second year.
2n K. Wightiana (Am. 1. c. ii. 546); foliis late ovalibus obtuse
acumiuatis basi cuneatis, tilamentis discretis, ovariis triovulatis,—Wight,
Cat. No. 2478.
HAB. In Zeylania, alt. 2-3000 ped., Walker!; Malabar, WigM.--
Frutex scaudens, glaber, cortice rugoso fusco. Ramuli abbreviation ^emper ?).
Folia basi cuueata, iutcgra vel vix denticulata, subtus pallida, 2-3 poll, longa, £-2
lata, petiolo J-poll, feduncuti axillares, validi, pctiolum viz superantes,- bracteis
pluribus squamui(brmib\is ovatis deciduis. Sapaia iusequalia, parva. Petala 9,
ovalia, obtusji, int. uiiuora. Bacca prioris.
This appears to be a more rigid shrub than the last, with smaller and broader
leaves, dud short, thick, woody branches. "We have not seen the nyxle plant. It is
worthy of note that K. Jayonico, is said by Siobold and 2uccarini to have also occa-
sionally three ovules and seeds.
3. K. scandeas (Blume? II. Jav. Scluz. p. 9.1.1).
A specimen of a Kadsum in the Bciithainiiui Herbarium, collected by Griffith at
Malacca, is very distinct from cither of the former species, and probably belongs to
£. scaMlew; but as it consists of a single leaf attached to the stem, and a few male
flowers, we do not feel justified in appending a description of that plant, especially
as Blurne's figure and description of the aiidraeiuui are unsatisfactory, aud also not
easily recoucilcd wtth what we see in the single flower -which we have been able to
examine. The leaves of K, scandeiis (and of kour plant) are ovate or ova to-oblong
and acumiaate, quite entire, glabrous, 4-6 inches long, and 2-4 broad, with a pe-
tiole 1-2 inches long. The Sowers ore axillary and solitary, aud the pedicel is
shorter than the petiole. Blume further describes the stamens as free on the cylin-
drical toro^, with the connective extending beyond the anther into a Heshy gibbous
process, This docs not seem to be the case in the specimen, from Malacca, bat the
liower has been so much compressed tliat we cannot determine the structure with
anything like accuracy. K. sctmdens is further readily recognizable by the shape of
the carpels, which arc terminated by an obtuse hooked mucro.
2. SPHJEROSTEMA, Bl.
Sepala 3. Petala 6-9. Stamina 15 vel plura, monadelpha. C&r-
pella globosa, sccus toruin cylmdricura splcata.—Frutices tcandtMtw,
volumes, gkberrimi, floribus albisjlavidis vel ntbescmtibus.
\ 1. Filawentis liasi mwiadetyUs, apice liberis.
\. S. grandiflorum (BL JP1. Jav. Schiz. 17); foliia ovato- vel ob-